Differential Scatterometry for Critical Dimension Metrology
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current scatterometry-based CD metrology faces challenges such as cross-talk of floating parameters, variation in optical properties, large target size, long calculation times, and time-consuming setup recipes, particularly in advanced lithographic processes where high-quality CD measurements are critical for yield improvement and control.
Innovation Solution
A method involving illumination of periodic targets with different critical dimension biases, measuring scattered radiation intensities, and determining a differential signal to accurately determine critical-dimension-related properties, which approximates to zero at a 1:1 line-to-space ratio, enabling precise CD measurements and dose control without the need for complex models or large targets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If conventional scatterometry is used with large targets (40 μm by 40 μm), then measurement is simplified and targets can be positioned in scribe lane, but target size is large and cannot be positioned amongst product features
Solution Approach 1:
The patent uses asymmetric target designs where the target structure itself exhibits asymmetry that is sensitive to overlay errors. By measuring the asymmetry of the target pattern after lithographic exposure, overlay accuracy can be determined without requiring large target areas or complex illumination schemes. The asymmetric target allows in-die measurements among product features while maintaining measurement capability.
Solution Approach 2:
The patent changes the parameter of target size from conventional large dimensions (40 μm) to smaller dimensions suitable for in-die placement. By combining smaller target size with asymmetric design and differential measurement techniques, the patent achieves both reduced target area and maintained measurement accuracy through differential signal processing that eliminates sensitivity to optical property variations.
2Measurement precision
If iterative reconstruction methods are used for CD determination, then CD measurement can be performed, but calculation time is long
Solution Approach 1:
The patent extracts only the essential measurement information needed for CD determination by using differential measurements that directly yield CD-related properties. Instead of performing full iterative reconstruction of the entire target structure, the method extracts specific differential signals that are directly proportional to CD variations, eliminating the need for time-consuming iterative calculations while maintaining measurement accuracy.
Solution Approach 2:
The patent replaces the mechanical iterative reconstruction process with a direct differential measurement approach. By measuring differential signals from targets with different CD biases and using the known relationship that the differential signal approximates zero at 1:1 line-to-space ratio, the method directly calculates CD without requiring iterative mathematical reconstruction, thereby dramatically reducing computation time.
3Adaptability or versatility
If floating parameters are used in scatterometry models, then model flexibility is increased, but cross-talk between parameters occurs and measurement reliability decreases
Solution Approach 1:
The patent extracts and eliminates the problematic floating parameters from the measurement model by using differential measurements between targets with different CD biases. The differential approach removes sensitivity to variations in optical properties and other floating parameters, leaving only the CD-related signal that is directly measurable and reliable, thereby eliminating cross-talk while maintaining the ability to measure CD accurately.
4Manufacturing precision
If optical property variations are present in the wafer, then process variations occur, but conventional scatterometry is sensitive to these variations causing measurement errors
Solution Approach 1:
The patent uses asymmetric target designs and differential measurement schemes that are inherently insensitive to optical property variations. By measuring the asymmetry of the target pattern rather than absolute intensity values, the method eliminates sensitivity to changes in optical properties across the wafer, allowing accurate CD measurements even in the presence of process variations and optical property gradients.
Solution Approach 2:
The patent implements a feedback mechanism where the differential signal from measurements is used to determine and correct for process variations. By continuously measuring the differential signal and comparing it to the expected zero value at 1:1 line-to-space ratio, the system can detect and compensate for optical property variations, maintaining measurement accuracy across different wafer conditions.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach provides accurate, fast, and model-independent CD measurements and dose control, is insensitive to optical property variations, and allows for smaller target sizes, reducing calculation time and improving in-die capability.
Implementation Method 1
measuring respective intensities of radiation scattered by the at least two targets
Data Source
AI summary
A method of determining a critical-dimension-related property, such as critical dimension (CD) or exposure dose, includes illuminating each of a plurality of periodic targets having different respective critical dimension biases, measuring intensity of radiation scattered by the targets, recognizing and extracting each grating from the image, determining a differential signal, and determining the CD-related property based on the differential signal, the CD biases and knowledge that the differential signal approximates to zero at a 1:1 line-to-space ratio of such periodic targets. Use of the determined CD-related property to control a lithography apparatus in lithographic processing of subsequent substrates. In order to use just two CD biases, a calibration may use measurements on a “golden wafer” (i.e. a reference substrate) to determine the intensity gradient for each of the CD pairs, with known CDs. Alternatively, the calibration can be based upon simulation of the sensitivity of intensity gradient to CD.